-R8 will set current bias thru the diff pair jfet input stage; Voltage across R8 divided by R8 resistance should have twice the current the jfets were matched at; Putting a trimmer pot on R8 can get you to this value.
-R27 sets the Vgs + VRs of the Aleph Active CS mosfet; Active because of R24 and C3; R24 will adjust the Aleph ACS Gain -presently at 50% with 1.2k I think; It's been awhile since Alephs were the DIYers fashion statement (2003-2005?) but do a search and you'll find how the math or actual measurements were done. What the combination of R24 and C3 does is to provide feedback to the CS, thereby making it Active, and asks for more current if the load requires more. Think of loads with impedance dipping instantaneously with music's varying frequencies. This sets it apart from a Constant Current Source(CCS)
-R7 adjusts output offset to zero by matching the voltage Vgs+VRs of the output power mosfets (Q7, Q8) to the Vgs+VRs of the ACS mosfets (Q5, Q6) adjusted by R27. Here you can see the importance of matching Vgs of the power mosfets and also the Source resistors.
Three questions before I start to populate the board:
1. Can Q2, Q3 and Q4 be put in in either direction?
2. The build guide suggests that you set R27 to approximately 68K when you install the pot. Exactly how do you do that? When I measure w104 across the outside pins I get 103K; with 50K across the left pin to the center and 53K from center to right pin. 3 turns counter-clockwise leaves the 103K the same but the two other measures go to 38K and 65K. So what is the process of setting the pot to 68K at installation?
3. Henry notes above some difficulty in getting the solder to adhere to pad. He advised to up the heat and play with the solder to get more flux on the pad. Anyone else have a similar problem and suggested solution?
1. Can Q2, Q3 and Q4 be put in in either direction?
2. The build guide suggests that you set R27 to approximately 68K when you install the pot. Exactly how do you do that? When I measure w104 across the outside pins I get 103K; with 50K across the left pin to the center and 53K from center to right pin. 3 turns counter-clockwise leaves the 103K the same but the two other measures go to 38K and 65K. So what is the process of setting the pot to 68K at installation?
3. Henry notes above some difficulty in getting the solder to adhere to pad. He advised to up the heat and play with the solder to get more flux on the pad. Anyone else have a similar problem and suggested solution?
1. No. Make sure you have the correct direction for the pins (C,B,E) to match your device. Check the data sheets for your device.
3. Note that the correct remedy is some flux on the pad prior to soldering if one has trouble for the solder to adhere. I am just too lazy to drive all the way in to the city to buy some flux, so I use a "workaround" like us software devs call it.
Notice on the board there are pads for a fixed R27 across the pot? Those make good test points to measure the setting before powering up.
I'm slowly accumulating parts for this build. For those that have built this, is there a noticeable thump at power-down? Has anyone used the "soft start & speaker turn-on delay" boards from the DIY store with this build or is it not needed? I see that there is a preorder for these boards in the store but was wondering if they're truly needed. Thoughts?
I am going to run the soft start board from the DIYAudio store in the Aleph J build that I am in the middle of getting started with. I don't think the soft start board will do much with the turn off thump, as I think that's residual capacitor discharge (not like I know though). The soft start board is more about lessening the strain of the inrush current during the power on cycle, letting transformer settle down before starting the capacitor charging.
This weekend's task (I am at the beach) is sorting the components into baggies with cards showing the part designation (C1, R1, etc.) I have a few boxes from Mouser, and some assorted other sources, I made up the index cards from the BOMs, and I will sort them into the correct baggies this weekend. Over the next week/weekend, I will confirm the parts and do some matching with the multi-meter. At some point next weekend, or so, I will start board stuffing.
I have the 5u deluxe chassis from the store, I will be drilling holes to mount Neutrik RCA and XLR jacks before putting it together
Perhaps later in October, I can tell you if there is a turn off thump with the soft start board in place.
This weekend's task (I am at the beach) is sorting the components into baggies with cards showing the part designation (C1, R1, etc.) I have a few boxes from Mouser, and some assorted other sources, I made up the index cards from the BOMs, and I will sort them into the correct baggies this weekend. Over the next week/weekend, I will confirm the parts and do some matching with the multi-meter. At some point next weekend, or so, I will start board stuffing.
I have the 5u deluxe chassis from the store, I will be drilling holes to mount Neutrik RCA and XLR jacks before putting it together
Perhaps later in October, I can tell you if there is a turn off thump with the soft start board in place.
Thanks ZM!
Powered up one channel and every thing looks nice... voltage across R8 is around 8.43 volts, around R7 around 4.23 Volts. Voltage on the sources of the 2SJ108's around +0.12 volts,
Voltage around the 0.47 ohm MOSFET's source resistors are around 0.6 Volts. Thanks Buzz for good matching! So biased pretty high. 30 Watts/ Mosfet.
Offset is around 0.014 volts, but with R7 easily trimmed to 0.00 volt.
Only the R27 variable resistor doesn't make much difference for the bias.. turning 3 a 4 turns doesn't make much difference... does it have to be a lower resistance to have less bias?
Is there an explanation of the circuit, Aleph J current source, like you did for the F5?
Must get some sleep now....
Thanks, Walter
I ended up with 21k for R27 so keep turning. those 25 turn pots just keep goin round and round
Do you have a jumper in the "other" R27? (Between the caps)
If not, your Aleph CCS is running wide open, and the pot isn't connected to the circuit.